|
benchmarks
|
add no-op network call, and a simple benchmark that exercises it
|
2010-08-09 20:21:41 +00:00 |
|
sherpa
|
Initial checkin of sherpa backend. These files probably don't belong here in the long term, but this is expedient.
|
2010-02-26 18:52:25 +00:00 |
|
test
|
cleaned up some code, moved pretty_print into its own thread, and fixed some buggy stats computations
|
2010-08-23 23:28:29 +00:00 |
|
util
|
added shutdown utility
|
2010-08-16 22:11:19 +00:00 |
|
adriana-lima.awk
|
initial import; removed cruft from mert's tarball, tweaked make's clean targets
|
2010-01-23 02:13:59 +00:00 |
|
CMakeLists.txt
|
add new server implementation that uses FILE to buffer I/O, and handles a smaller, fixed number of connections -- leads to an almost 4x speedup on no-ops on nehalem
|
2010-08-11 22:46:55 +00:00 |
|
cmds.txt
|
initial import; removed cruft from mert's tarball, tweaked make's clean targets
|
2010-01-23 02:13:59 +00:00 |
|
datapage.cpp
|
fix latch/pin leak
|
2010-08-12 01:55:30 +00:00 |
|
datapage.h
|
Implemented snowshoveling, with caveats.
|
2010-08-05 17:43:46 +00:00 |
|
datatuple.h
|
Generalized comparator, fixed some bugs in record length handling.
|
2010-02-23 17:05:47 +00:00 |
|
diskTreeComponent.cpp
|
scattered, and poorly tested fixes. finer grained latching for tree nodes; throttle c0-c1 merge when c0 is small; fix delta computations for c0 throttling; disable short pauses for c1-c2 merge. this gets the bulk load throughput up to 8500, but workloada is still slow
|
2010-08-21 03:09:18 +00:00 |
|
diskTreeComponent.h
|
added destructor to diskTreeComponent (fixes file handle / memory leak)
|
2010-08-05 17:39:31 +00:00 |
|
install.sh
|
install.sh script for yroots
|
2010-02-21 17:04:30 +00:00 |
|
logserver.cpp
|
add new server implementation that uses FILE to buffer I/O, and handles a smaller, fixed number of connections -- leads to an almost 4x speedup on no-ops on nehalem
|
2010-08-11 22:46:55 +00:00 |
|
logserver.h
|
Refactoring of logtable. Cleaned a bunch of includes, and logtable is now a template. This is in preparation for the new merge policy
|
2010-03-17 21:51:26 +00:00 |
|
logstore.cpp
|
remove old comments
|
2010-08-24 00:40:48 +00:00 |
|
logstore.h
|
scattered, and poorly tested fixes. finer grained latching for tree nodes; throttle c0-c1 merge when c0 is small; fix delta computations for c0 throttling; disable short pauses for c1-c2 merge. this gets the bulk load throughput up to 8500, but workloada is still slow
|
2010-08-21 03:09:18 +00:00 |
|
memTreeComponent.cpp
|
Refactoring of logtable. Cleaned a bunch of includes, and logtable is now a template. This is in preparation for the new merge policy
|
2010-03-17 21:51:26 +00:00 |
|
memTreeComponent.h
|
remove old comments
|
2010-08-24 00:40:48 +00:00 |
|
mergeManager.cpp
|
remove old comments
|
2010-08-24 00:40:48 +00:00 |
|
mergeManager.h
|
remove old comments
|
2010-08-24 00:40:48 +00:00 |
|
merger.cpp
|
remove old comments
|
2010-08-24 00:40:48 +00:00 |
|
merger.h
|
two unrelated changes: release rb_mut before getting header_mut on read path (breaks delta tuples), and implement first-cut support for pulling in statistics when reopening logstores (required for correct recovery)
|
2010-06-21 20:03:05 +00:00 |
|
mergeStats.cpp
|
Rewrite merge thread synchronization code, update status messages, implement preliminary (and ineffective) admission control, and force write merged data every megabyte (so that prograess can be tracked by admission control. Also, refactor quite a few things.
|
2010-05-19 23:42:06 +00:00 |
|
mergeStats.h
|
scattered, and poorly tested fixes. finer grained latching for tree nodes; throttle c0-c1 merge when c0 is small; fix delta computations for c0 throttling; disable short pauses for c1-c2 merge. this gets the bulk load throughput up to 8500, but workloada is still slow
|
2010-08-21 03:09:18 +00:00 |
|
network.h
|
fix a deadlock (double setting thread_fd = -1 in race at startup), lazily spawn new worker threads. This will save some ram, and will lead to shorter server thread listings in gdb
|
2010-08-11 23:27:39 +00:00 |
|
newserver.cpp
|
add memTreeComponent::batchedRevalidatingIterator; amortize cost of red black latching during merge
|
2010-08-18 17:29:25 +00:00 |
|
NOTES
|
initial import; removed cruft from mert's tarball, tweaked make's clean targets
|
2010-01-23 02:13:59 +00:00 |
|
regionAllocator.h
|
Implemented snowshoveling, with caveats.
|
2010-08-05 17:43:46 +00:00 |
|
requestDispatch.cpp
|
re-enable old merge policy for c1-c2 merger; implement shutdown (sort of)
|
2010-08-17 21:23:39 +00:00 |
|
requestDispatch.h
|
add new server implementation that uses FILE to buffer I/O, and handles a smaller, fixed number of connections -- leads to an almost 4x speedup on no-ops on nehalem
|
2010-08-11 22:46:55 +00:00 |
|
server.cpp
|
incease UPDATE_PROGRESS_DELTA by 10x, and FORCE_INTERVAL by 2.5x; tick no longer grabs a writelock on header_mut, and instead uses its own mutex
|
2010-06-21 22:59:05 +00:00 |
|
simpleServer.cpp
|
re-enable old merge policy for c1-c2 merger; implement shutdown (sort of)
|
2010-08-17 21:23:39 +00:00 |
|
simpleServer.h
|
re-enable old merge policy for c1-c2 merger; implement shutdown (sort of)
|
2010-08-17 21:23:39 +00:00 |
|
tcpclient.cpp
|
fix a deadlock (double setting thread_fd = -1 in race at startup), lazily spawn new worker threads. This will save some ram, and will lead to shorter server thread listings in gdb
|
2010-08-11 23:27:39 +00:00 |
|
tcpclient.h
|
Hacky histogram generation; API's for client-side iterators.
|
2010-03-03 23:20:05 +00:00 |
|
tuplemerger.cpp
|
cleanup statistics gathering and reporting.
|
2010-03-05 19:07:47 +00:00 |
|
tuplemerger.h
|
cleanup statistics gathering and reporting.
|
2010-03-05 19:07:47 +00:00 |
960ff041e2